Re: [xml-dev] 2007 Wish List - was RE: [xml-dev] 2007 Predictions

["Pete Cordell" <petexmldev@tech-know-ware.com>]

Commenting on a few responses in one...

I agree with Michael Champion that it would have to be a one-to-one mapping. 
I also agree with Michael Kay that it would have to be endorsed by W3C (e.g. 
XSD part 3).

I don't really agree that the wide support for XSD in tooling makes it 
unsuitable for a compact format.  All it needs is one implementation on the 
web that does compact format to regular format and vice versa and you're 
done (similar to what XSV does for validation).  This could be used by 
humans, or even applications.  (For example, in a basic scenario a developer 
could develop in compact format, then use a Perl script or similar to see 
what regular format files were out of date and use the web service to 
convert the latest compact form to the regular form.)

Plus, some environments may be sufficiently controlled that they know they 
don't need the regular format.

And that's worse case.  It's really just a different front end, and I could 
imagine tool developers could integrate such a thing quite quickly if there 
was suitable customer demand.

Pete

---Original Message From: "Michael Champion"
> > Original Message From: Pete Cordell > >
> > Any prospect for a compact notation for XML schema in a similar vain to
> > Compact Relax NG?
>
> There is http://dret.net/projects/xscs/ and I suspect there are others.
> This > came to my attention because it is used by http://brainml.org/
>
> >
> > Would this be helpful?
>
> Probably.  As far as we can tell, most XML schemas are written by
> developers (See Stan's writeup of the survey conducted by the Michigan
> Business School) http://2006.xmlconference.org/proceedings/34/slides.pdf)
> and it stands to reason that they would prefer working with a more code-
> like syntax.
>
> Likely?  I doubt it, partly for the reason David Megginson noted: The
> main thing XSD has going for it is ubiquity. Unless there is a perfect
> two-way mapping between some XSD-CS and XSD-XML, and there are tools on
> every platform to do the translation, who will *really* find it easier to
> work with the compact syntax  in the real world?  Or to put it
> differently, is the ugliness of XSD-XML really enough of a pain point to
> justify tool vendors to support the compact syntax?  I'd guess the
> complexity of the "schema object model" is more of a deterrent, so visual
> tools that help guide people through the abstract model will be more
> helpful than syntax tools. (The popularity of RNG compact syntax may not
> be relevant here because RNG does *not* have such a complex underlying
> model and the XML syntax really is a pain point ... anyone have thoughts
> on this?).
>
--
=============================================
Pete Cordell
Tech-Know-Ware Ltd
for XML to C++ data binding visit
http://www.tech-know-ware.com/lmx
(or http://www.xml2cpp.com)
=============================================